Difficulty understanding speech in noisy conditions is the primary complaint of people with hearing impairment, and is also known to be problematic for individuals listening to a non-native language. The long-term goal of the proposed research is to understand the role of linguistic factors (e.g. listener language experience, linguistic content of signals and noise) in modulating the intelligibility of speech in noisy conditions. The specific aims of this application are to establish the relationship between long-term language experience and speech understanding in speech noise, and to develop a training protocol to improve speech intelligibility in noisy conditions. The rationale for this research is that it will provide crucial linguistic insight into the general phenomenon of speech-in-noise understanding and facilitate the improvement of assessment and rehabilitative techniques for mono- and multi-lingual clinical hearing populations. These aims will be achieved by assessing sentence intelligibility in speech noise over a range of signal-tonoise ratios for three groups of listeners: monolingual English speakers, (early) bilingual speakers of Mandarin and English, and non-native (late bilingual) speakers of English whose native language is Mandarin. Listeners will be tested on speech targets in either English or Mandarin, as well as with interfering speech noise in each language. Training conditions will be assessed in a series of speech-in-noise training experiments with monolingual English-speaking young adults.